Hi Ravi -
The shutdown / in-memory corruption essentially means that something
stomped on an in-memory structure, and when this was detected, the
filesystem shut down. In the past, we have seen LVM+XFS cause stack
overflows, which could maybe cause this, but would more likely oops.
In general, these are pretty hard to debug. :(
What version of LVM are you using?
The output of xfs_repair might also be helpful. (I think force_shutdown
causes a dirty log; you'd need to mount the filesystem to replay the log
first, _then_ run xfs_repair to see if there are problems. xfs_repair
on a dirty log will generate lots of errors and generally should not be
done - see the -L option).
-Eric
On Tue, 30 Apr 2002, Ravi Wijayaratne wrote:
> Hi all
>
> We are seeing the following problem after some heavy
> I/O on our Software Raid 0 with LVM XFS system.
>
> xfs_force_shutdown(lvm(58,0),0x8) called from line
> 1039 of file xfs_trans.c. Return address = 0xc01cf019
> kernel Apr 26 10:54:55
> Corruption of in-memory data detected. Shutting
> down filesystem: lvm(58,0) kernel Apr 26 10:54:55
>
> When we run xfsrepair afterwards we get many messages
> about file system integrity violations.
>
> We are running Linux 2.4.18 kernel. I have seen
> several postings about this problem in the xfs mailing
> list and the FAQ. Is there a fix for the above problem
> ?
--
Eric Sandeen XFS for Linux http://oss.sgi.com/projects/xfs
sandeen@xxxxxxx SGI, Inc.
|